Strategic Wishful Thinking: Implications for Forecasts

Survey forecasts depart from rational expectations in ways that look contradictory: consensus forecasts under-react to news, individual forecasts of macroeconomic variables tend to over-react, and individual forecasts of financial variables such as earnings per share under-react. This paper offers a single mechanism that produces all three. Forecasters are uncertain about how precise a public signal is, and exhibit “wishful thinking” — they choose a subjective interpretation of that precision to maximise anticipatory utility, subject to a penalty from taking suboptimal actions — within the quadratic-Gaussian coordination framework of Angeletos and Pavan (2007), where a forecaster’s payoff depends on her own accuracy, on an incentive to herd toward or stand out from the consensus (ρ), and on a separate externality (γ) from how accurate the consensus itself is. Two channels compete: a fundamental-uncertainty channel pushing every forecaster to overestimate the public signal’s precision, and an aggregate-error channel whose sign depends on the combination of ρ and γ. When γ is sufficiently high the channels reinforce and all individual forecasts over-react; when γ is sufficiently low the aggregate-error channel dominates and all under-react; and for intermediate γ neither dominates for everyone, so the unique equilibrium is in mixed strategies and ex-ante symmetric forecasters endogenously “agree to disagree” — some underestimating the public signal’s precision while others overestimate it. Because γ is argued to be negative in financial markets, where others’ mistakes represent profitable trading opportunities, this maps the over-reaction/under-reaction split onto macroeconomic versus financial forecasts; the model further predicts dispersion in individual Coibion–Gorodnichenko coefficients — zero under rational expectations — and that endogenous disagreement is more likely when public information is imprecise, which the paper notes is broadly consistent with evidence that disagreement about public information rises in downturns. The paper is theoretical: it derives testable predictions and relates them to existing empirical findings rather than estimating the model.

Naveen Gondhi | Macro Paper Warehouse

Strategic Wishful Thinking: Implications for Forecasts